[FFmpeg-devel] [PATCH 0/2] colorkey vf, int math conversion, libavfilter questions

Timo Rothenpieler timo at rothenpieler.org
Tue Apr 28 19:31:31 CEST 2015

Still not fully intended for merging, as there are a few issues/questions left.

First thing is the conversion of the float based algorithm to integer math.
The code i came up with works, i can't spot a visual difference. But i have no
idea if this is the most optimal way to approach this.
I left the conversion to int math in a seperate commit for now, for easier
review of those changes.

The other issue is the pixel format conversions done by this filter.
It seems like libavfilter expects a filter to be able to convert from any of
its input formats to any of its output formats.
This isn't the case for this filter. The only "conversion" it does is from
BGR0 to BGRA, 0RGB to ARGB, and so on, if the input format doesn't already
contain an alpha channel.
I haven't found any way to do such a mapping in query_formats, what would be 
the correct way to handle this?

Timo Rothenpieler (2):
  avfilter/vf_colorkey: Add colorkey video filter
  avfilter/vf_colorkey: Transform calculations to integer math

 Changelog                 |   1 +
 MAINTAINERS               |   1 +
 doc/filters.texi          |  39 +++++++
 libavfilter/Makefile      |   1 +
 libavfilter/allfilters.c  |   1 +
 libavfilter/vf_colorkey.c | 282 ++++++++++++++++++++++++++++++++++++++++++++++
 6 files changed, 325 insertions(+)
 create mode 100644 libavfilter/vf_colorkey.c


More information about the ffmpeg-devel mailing list